Plant-associate interactions and diversification across trophic levels

Post by Jeremy Yoder

A new study pub­lished in Evol­u­tion Let­ters invest­ig­ates how host plants and their asso­ci­ates coe­volve. Author Jeremy Yoder tells us more:

Spe­cies that live in close asso­ci­ation with anoth­er organ­ism form some of the most diverse groups of liv­ing things we know, things like but­ter­flies, beetles, para­sit­ic worms, and para­sit­oid wasps. Evol­u­tion­ary bio­lo­gists have long sus­pec­ted that intim­ate coe­volu­tion of asso­ci­ates with their hosts — adapt­a­tion of the asso­ci­ate to the host, and the host to the asso­ci­ate — helped drive this diversification.

It’s chal­len­ging to catch coe­volu­tion­ary spe­ci­ation in pro­gress. We can look for cases in which dif­fer­ent pop­u­la­tions of an asso­ci­ate spe­cies have star­ted using dif­fer­ent hosts to test wheth­er adapt­ing to those dif­fer­ent hosts cre­ates genet­ic dif­fer­en­ti­ation between the asso­ci­ate pop­u­la­tions. Even that study design doesn’t quite cap­ture the res­ults of pair­wise inter­ac­tion between a single asso­ci­ate and its host, though.

As bio­lo­gists who study coe­volu­tion, we recog­nized that a par­tic­u­lar sig­nal of host-asso­ci­ate coe­volu­tion could be cap­tured in genet­ic data for a host and its asso­ci­ate. When pop­u­la­tions of a spe­cies become adap­ted to loc­al con­di­tions in dif­fer­ent hab­it­ats they encounter, they often show a pat­tern of eco­lo­gic­al isol­a­tion, in which pop­u­la­tions adap­ted to dif­fer­ent hab­it­ats are more genet­ic­ally dif­fer­ent from each oth­er than we would oth­er­wise expect. For intim­ate asso­ci­ate spe­cies, the most import­ant hab­it­at is often their hosts — so, poten­tially, they might show pop­u­la­tion genet­ic dif­fer­ences that par­al­lel genet­ic dif­fer­ences between their hosts’ pop­u­la­tions. A pat­tern of par­al­lel genet­ic dif­fer­ences in host and asso­ci­ate pop­u­la­tions would catch the pro­cess of coe­volu­tion­ary spe­ci­ation in its earli­est stages.

To test for this pat­tern, we needed genet­ic data from asso­ci­ate spe­cies and their hosts, sampled in the same set of loc­a­tions across their over­lap­ping home ranges. Col­lect­ing genet­ic data for not one but two spe­cies makes for a pretty com­plic­ated (and poten­tially expens­ive) research pro­ject, but we knew it could be done — part of my own Ph.D. dis­ser­ta­tion, pub­lished in 2013, included this kind of genet­ic data for Joshua trees (Yucca brevi­fo­lia and Y. jae­geri­ana) and their spe­cial­ized pol­lin­at­ors Tege­tic­ula syn­thet­ica and T. anti­thet­ica. We searched for pub­lished sci­entif­ic papers report­ing host-asso­ci­ate data sets, and even­tu­ally came up with 15 papers report­ing the kind of data we needed, from 20 pairs of asso­ci­ate spe­cies and their hosts, all of them plants. We focused on plants as hosts spe­cific­ally because there is a long his­tory of study­ing insect herb­i­vores’ inter­ac­tions with plants — and indeed most of the asso­ci­ate spe­cies were insects.

The sup­port­ing inform­a­tion files from the papers we’d found gave us a basis to re-exam­ine the 20 host-asso­ci­ate pairs using the same ana­lyt­ic frame­work, and then apply meta-ana­lys­is to look for gen­er­al pat­terns. Over­all, asso­ci­ate spe­cies’ pop­u­la­tion genet­ic dif­fer­ences par­alleled those of their host plants, to a degree bey­ond what could be explained by vari­ation in cli­mate among the sites sampled in each data set, or simple isol­a­tion by geo­graph­ic dis­tance. That’s the “pop­u­la­tion genet­ic fin­ger­print” of coe­volu­tion-driv­en diver­si­fic­a­tion we thought we might find.

We also sep­ar­ated the host-asso­ci­ate pairs out by the nature of their asso­ci­ation, to see wheth­er dif­fer­ent types of inter­ac­tion had dif­fer­ent diver­si­fy­ing effects. The­ory about the nature of mutu­al­ist­ic versus ant­ag­on­ist­ic spe­cies inter­ac­tions has gen­er­ally pre­dicted that ant­ag­on­ism should drive diver­si­fic­a­tion more than mutu­al­ism. Ant­ag­on­ists coe­volve with their hosts via arms races or fre­quency-depend­ent selec­tion arising from hosts’ immune responses, and these can send indi­vidu­al pop­u­la­tions off on their own loc­al coe­volu­tion­ary tra­ject­or­ies. Mutu­al­ists, on the oth­er hand, more often coe­volve to match with their hosts and main­tain the bene­fits of inter­ac­tion, which should tend to main­tain sim­il­ar­it­ies across pop­u­la­tions of both spe­cies. Across the data we com­piled, ant­ag­on­ist­ic asso­ci­ates like para­sites and herb­i­vores did indeed show some­what stronger par­al­lel genet­ic dif­fer­ences with their host plants than mutu­al­ists did. Mutu­al­ists also showed sig­ni­fic­ant isol­a­tion by geo­graph­ic dis­tance, ant­ag­on­ist­ic asso­ci­ates gen­er­ally did not — this lines up with what we expect if mutu­al­ists tend to track the geo­graph­ic dis­tri­bu­tions of their hosts.

The stud­ies we com­piled rep­res­ent a lim­ited look at the huge diversity of spe­cies that live in intim­ate asso­ci­ation with hosts. Still, we think our res­ults are excit­ing even as a pre­lim­in­ary answer to the ques­tion of how spe­cies inter­ac­tions can cre­ate new biod­iversity — and we’re hope­ful this paper will inspire the kind of data col­lec­tion needed to provide a more com­pre­hens­ive picture.

Jeremy Yoder is a pro­fess­or in the Depart­ment of Bio­logy at Cali­for­nia State Uni­ver­sity North­ridge. The ori­gin­al art­icle is freely avail­able to read and down­load from Evol­u­tion Letters.